64bit windows release
Moderator: Moderators
64bit windows release
While a 32bit version works, with 64bit versions you can make assumptions about the cpu such as enabling SSE (a 64bit cpu without sse?!!?!?!?), so there should be a speedup as a result.
Re: 64bit windows release
Will it synch with 32bits release?
Re: 64bit windows release
Nice, not only cross operating system but also cross architecture compiling
(is there even a mingw64 configure target for GCC?)
EDIT: Also IMHO the speed difference between 32 and 64 bit is overrated, I didn't notice anything.

EDIT: Also IMHO the speed difference between 32 and 64 bit is overrated, I didn't notice anything.
-
- Posts: 933
- Joined: 27 Feb 2006, 02:04
Re: 64bit windows release
There isn't too much of a speed difference for average applications, but you get the ability to address much more RAM in your process and you could just assume all the CPUs have SSE and SSE2. SSE2 is supported on all Athlon64s, and Intel chips that can do 64bit, and would probably speed up some calculations a LOT on Netburst architecture Intel processors.
Re: 64bit windows release
It might make some bugs easier to find too. Currently there seem to be a few serious issues that plague 64 bit linux but not 32, (eg the KAIK that shipped with 0.76 freezing immediately). I'd be willing to bet that some of my crashes (where no one else crashes) are related to 64 bit issues as well.
Re: 64bit windows release
I agree, the benefits of 64bit itself are not from movign to 64bit, but by being able to assume support for SSE and SSE2 instruction sets.
I hadnt thought about the testing potential for fixing 64bit errors.
I'd assume tobi is busy with things, as am I, can someone look into a mingw64 configure target for GCC?
I hadnt thought about the testing potential for fixing 64bit errors.
I'd assume tobi is busy with things, as am I, can someone look into a mingw64 configure target for GCC?